| Peer-Reviewed

Using Machine Learning for the Development of a Maintenance Management System: Case Study of Kenya

Received: 31 May 2022     Accepted: 29 June 2022     Published: 12 July 2022
Views:       Downloads:
Abstract

A Maintenance Management System (MMS) was first developed in the 1982 for implementation in the Arizona Department of Transportation in the United States. It allows for a forecast of future maintenance activities for a road network which deteriorates over time. Successive enhancements to the original MMS have been made over the years by different researchers, including some by the first author. The primary enhancements have been in the formulation and solution algorithms. The initial solution algorithms were Linear Programming (LP) and Dynamic Programming (DP), which, in some previous works, were replaced by genetic algorithms due to their efficiency over LP and DP. In this paper, we propose a Machine Learning (ML) framework for the development of a MMS, which can be a better approach than previously developed approaches. The ML framework uses a Python-based solution methodology in conjunction with geo-spatial modeling, which appears more attractive and efficient in working directly with GIS maps and databases. With respect to application, the attention is focused on African countries using Kenya as a case study example. A recent report on state of Kenyan roads found over 35 percent of Kenyan roads to be still in poor condition even though a comparison of the condition of the roads between 2003 and 2018 showed a successive improvement in road condition over the years. Poor road condition affects mobility and, in turn affects the country’s economy. We adopt a Markov Decision Process to predict the maintenance actions to be undertaken for the Kenyan road network in order to keep an acceptable level of service quality over a specified planning horizon. A budget can then be estimated based on the cost of maintenance actions. A case study using Geographic Information System maps and databases demonstrates the effectiveness of the approach. The result shows that an MMS for Kenyan roads can help forecast the maintenance activities to be undertaken over a planning horizon. For more realistic practical applications, using some of our previous works as a guide, an algorithm to decide on the level of deterioration over time can be developed in future works which could consider factors like weather, vehicle mix, and traffic load.

Published in International Journal of Economics, Finance and Management Sciences (Volume 10, Issue 4)
DOI 10.11648/j.ijefm.20221004.12
Page(s) 166-172
Creative Commons

This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.

Copyright

Copyright © The Author(s), 2022. Published by Science Publishing Group

Keywords

Maintenance Management System, Markov Decision Process, Machine Learning, Road Safety and Mobility, Kenyan Roads

References
[1] Abdullah, J. (2007). Models for the Effective Maintenance of Roadside Features, Doctoral Dissertation, Morgan State University, Baltimore, MD.
[2] Cheu, R. L., Wang, Y., Fwa, T. F. Genetic Algorithm – Simulation Methodology for Pavement Maintenance Scheduling. Computer Aided Civil and Infrastructure Engineering, 19 (6), 446-455, 2004.
[3] Durango, P. L., Madanat, S. M. (2002). Optimal Maintenance and Repair Policies in Infrastructure Management under Uncertain Facility Deterioration Rates, An Adaptive Control Approach, Transportation Research, Part A, 36, 763-778.
[4] Golabi, K., Kulkarni, R. B., and Way, G. B. (1982). A Statewide Pavement Management System, Interfaces, 12 (6), 5-21.
[5] Jha, M. K. and Ogallo, H. (2021). Studying the Dynamic Sight Distance Problem with a Machine Learning Algorithm, presented at the 2021 Annual TRB Meeting, Washington, D.C., Paper Number: TRBAM-21-03783.
[6] Jha, M. K., H. Ogallo, and O. Owolabi (2014). A Quantitative Analysis of Sustainability and Green Transportation Initiatives in Highway Design and Maintenance, Procedia - Social and Behavioral Sciences 111, 1185 – 1194, Elsevier.
[7] Jha, M. K., S. Shariat, J. Abdullah, and B. Devkota (2012). Maximizing resource effectiveness of highway infrastrucrre maintenance inspection and scheduling for efficient city logistics operations, Procedia-Social and Behavioral Sciences, 49, 831-844, Elsevier.
[8] Jha, M. K., K. Kepaptsoglou, M. Karlaftis, and G. A. K. Karri (2011). A Bilevel Optimization Model For Large Scale Highway Infrastructure Maintenance Inspection and Scheduling Following a Seismic Event, in Computational Methods in Earthquake Engineering, M. Papadrakakis, M. Fragiadakis, N. D. Lagaros (eds.), Volume 21, 515-526, Springer.
[9] Jha, M. K. (2010). Optimal Highway Infrastructure Maintenance Scheduling Considering Deterministic and Stochastic Aspects of Deterioration, in Sustainable and Resilient Critical Infrastructure Systems, K. Gopalakrishnan and S. Peeta (eds.), 231-248, Springer.
[10] Jha, M. K., S, Chacha, F. Udenta, and J. Abdullah (2010). Formulation and Solution Algorithms for Highway Infrastructure Maintenance Optimization with Work-Shift and Overtime Limit Constraints, Procedia Social and Behavioral Sciences 2, 6323-6331, Elsevier.
[11] Jha, M. K., B. Devkota, and B. Kattel (2010). An Optimization Approach to Maximize Service Quality subject to a given Resource Level for Highway Infrastructure Maintenance, in Industrial Resource Utilization and Productivity, A. Mital and A. Pennathur (eds.), Momentum Press.
[12] Jha, M. K. and J. Abdullah, (2008). A Probabilistic Bi-Level Optimization Approach to Highway Infrastructure Maintenance in Urban Areas, in Recent Advances in City Logistics: Proceedings of the 5th International Conference on City Logistics, E. Taniguchi and R. Thompson (eds.), Nova Science Publishers.
[13] Jha, M. K., Udenta, F., Chacha, S., and Karri, G. (2008). A Modified Arc Routing Problem for Highway Feature Inspection Considering Work-Shift and Overtime Limit Constraints, in New Aspects of Urban Planning and Transportation, World Scientific and Engineering Academy and Society (WSEAS) Press, pp. 105-109.
[14] Jha, M. K., K. Kepaptsoglou, M. Karlaftis, and J. Abdullah (2008). A BiLevel Optimization Model for Large-Scale Highway Infrastructure Maintenance Inspection and Scheduling, 10th International Conference on Application of Advanced technologies in Transportation, Athens, Greece, May 2008.
[15] Jha, M. K., K. Kepaptsoglou, M. Karlaftis, and J. Abdullah (2006). A Genetic Algorithms-Based Decision Support System for Transportation Infrastructure Management in Urban Areas, in Recent Advances in City Logistics: Proceedings of the 4th International Conference on City Logistics, pp. 509-523, E. Taniguchi and R. Thompson (eds.), Elsevier Publishing Company, Hardbound, ISBN: 0-08-044799-6, 554 pp.
[16] Jha, M. K., and Abdullah, J. (2006). A Markovian Approach for Optimizing Highway Life-Cycle with Genetic Algorithms by Considering Maintenance of Roadside Appurtenances. Journal of the Franklin Institute, 343, 404-419.
[17] Jha, M. K. and J. Abdullah (2006). A Probabilistic Approach To Maintenance Repair And Rehabilitation (MR&R) of Roadside Features, proceedings of the 14th Pan-American Conference on Traffic and Transportation Engineering (PANAM XIV), Las Palmas de Gran Canaria, Spain.
[18] Jha, M. K., D. Dave, and J. Abdullah (2004). An Integrated Approach to Highway Infrastructure Maintenance, proceedings of the 10th World Conference on Transport Research (WCTR), Istanbul, Turkey, July 2004.
[19] Jha, M. K., Abdullah, J., and Dave, D. (2004). GIS Application in Developing a Roadway Feature Inventory Program. Proceedings of the 2004 ESRI International User Conference, Paper Number 2128, San Diego, CA.
[20] Jha, M. K., J. Abdullah, and D. Dave (2004). Innovative Highway Maintenance with Dynamic Segmentation and the Markov Decision Process, proceedings of the XIII PanAmerican Conference, CD ROM Paper Number IS-17, Albany, NY.
[21] Jha, M. K. and Schultz. L. (2003). Development of an Integrated Highway Maintenance Management System: The Maryland Experience, presented at the AASHTO/TRB Maintenance Management Conference, Duluth, Minnesota, Transportation Research Circular, Number E-C052, 108-129, Transportation Research Board, Washington, D.C.
[22] Jha, M. K. and P. Schonfeld (2000). Integrating Genetic Algorithms and GIS to Optimize Highway Alignments. Transportation Research Record, Journal of the Transportation Research Board, 1719, 233-240.
[23] Kenya Roads Board: State of Our Roads 2018 (2018). Summary Report on Road Inventory and Conditions Survey Results and Policy Recommendations, Kenya Roads Board.
[24] Madanat, S., Ben-Akiva, M. (1994). Optimal Inspection and Repair Policies for Infrastructure Facilities. Transportation Science, 28 (1), 55-62.
[25] Madanat, S. (1991). Optimizing Sequential Decisions under Measurement and Forecasting Uncertainty: Application to Infrastructure Inspection, Maintenance and Rehabilitation. Ph.D. Dissertation, Massachusetts Institute of Technology, Cambridge, MA.
[26] Maji, A., and Jha, M. K. (2008), Maintenance Schedule of Highway Infrastructure Elements Using a Genetic Algorithm, Proc. of the International Conference on Civil Engineering in the New Millennium: Opportunities and Challenges (CENeM-2007), 150- year anniversary conference at Bengal Engineering and Science University, Shibpur, India, Vol. III, pp. 2108-2114.
[27] Maji, A and Jha, M. K. (2007). Modelling Highway Infrastructural Maintenance Schedule with Budget Constraint, Transportation Research Record 1991, Journal of the Transportation Research Board, 19-26.
[28] Mdptoolbox (2021). https://pymdptoolbox.readthedocs.io/en/latest/ (accessed September 27, 2021).
[29] Stephanos, P. and Hedfi, A. (2002). Maryland State Highway Administration's Project Selection Process: Integrating Network and Project-Level Analysis, Transportation Research Record 1816, Journal of the Transportation Research Board, 16-25.
[30] Wikepedia (2021). Markov decision process, https://en.wikipedia.org/wiki/Markov_decision_process, (accessed September 27, 2021).
Cite This Article
  • APA Style

    Manoj Kumar Jha, Hellon G. Ogallo. (2022). Using Machine Learning for the Development of a Maintenance Management System: Case Study of Kenya. International Journal of Economics, Finance and Management Sciences, 10(4), 166-172. https://doi.org/10.11648/j.ijefm.20221004.12

    Copy | Download

    ACS Style

    Manoj Kumar Jha; Hellon G. Ogallo. Using Machine Learning for the Development of a Maintenance Management System: Case Study of Kenya. Int. J. Econ. Finance Manag. Sci. 2022, 10(4), 166-172. doi: 10.11648/j.ijefm.20221004.12

    Copy | Download

    AMA Style

    Manoj Kumar Jha, Hellon G. Ogallo. Using Machine Learning for the Development of a Maintenance Management System: Case Study of Kenya. Int J Econ Finance Manag Sci. 2022;10(4):166-172. doi: 10.11648/j.ijefm.20221004.12

    Copy | Download

  • @article{10.11648/j.ijefm.20221004.12,
      author = {Manoj Kumar Jha and Hellon G. Ogallo},
      title = {Using Machine Learning for the Development of a Maintenance Management System: Case Study of Kenya},
      journal = {International Journal of Economics, Finance and Management Sciences},
      volume = {10},
      number = {4},
      pages = {166-172},
      doi = {10.11648/j.ijefm.20221004.12},
      url = {https://doi.org/10.11648/j.ijefm.20221004.12},
      eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ijefm.20221004.12},
      abstract = {A Maintenance Management System (MMS) was first developed in the 1982 for implementation in the Arizona Department of Transportation in the United States. It allows for a forecast of future maintenance activities for a road network which deteriorates over time. Successive enhancements to the original MMS have been made over the years by different researchers, including some by the first author. The primary enhancements have been in the formulation and solution algorithms. The initial solution algorithms were Linear Programming (LP) and Dynamic Programming (DP), which, in some previous works, were replaced by genetic algorithms due to their efficiency over LP and DP. In this paper, we propose a Machine Learning (ML) framework for the development of a MMS, which can be a better approach than previously developed approaches. The ML framework uses a Python-based solution methodology in conjunction with geo-spatial modeling, which appears more attractive and efficient in working directly with GIS maps and databases. With respect to application, the attention is focused on African countries using Kenya as a case study example. A recent report on state of Kenyan roads found over 35 percent of Kenyan roads to be still in poor condition even though a comparison of the condition of the roads between 2003 and 2018 showed a successive improvement in road condition over the years. Poor road condition affects mobility and, in turn affects the country’s economy. We adopt a Markov Decision Process to predict the maintenance actions to be undertaken for the Kenyan road network in order to keep an acceptable level of service quality over a specified planning horizon. A budget can then be estimated based on the cost of maintenance actions. A case study using Geographic Information System maps and databases demonstrates the effectiveness of the approach. The result shows that an MMS for Kenyan roads can help forecast the maintenance activities to be undertaken over a planning horizon. For more realistic practical applications, using some of our previous works as a guide, an algorithm to decide on the level of deterioration over time can be developed in future works which could consider factors like weather, vehicle mix, and traffic load.},
     year = {2022}
    }
    

    Copy | Download

  • TY  - JOUR
    T1  - Using Machine Learning for the Development of a Maintenance Management System: Case Study of Kenya
    AU  - Manoj Kumar Jha
    AU  - Hellon G. Ogallo
    Y1  - 2022/07/12
    PY  - 2022
    N1  - https://doi.org/10.11648/j.ijefm.20221004.12
    DO  - 10.11648/j.ijefm.20221004.12
    T2  - International Journal of Economics, Finance and Management Sciences
    JF  - International Journal of Economics, Finance and Management Sciences
    JO  - International Journal of Economics, Finance and Management Sciences
    SP  - 166
    EP  - 172
    PB  - Science Publishing Group
    SN  - 2326-9561
    UR  - https://doi.org/10.11648/j.ijefm.20221004.12
    AB  - A Maintenance Management System (MMS) was first developed in the 1982 for implementation in the Arizona Department of Transportation in the United States. It allows for a forecast of future maintenance activities for a road network which deteriorates over time. Successive enhancements to the original MMS have been made over the years by different researchers, including some by the first author. The primary enhancements have been in the formulation and solution algorithms. The initial solution algorithms were Linear Programming (LP) and Dynamic Programming (DP), which, in some previous works, were replaced by genetic algorithms due to their efficiency over LP and DP. In this paper, we propose a Machine Learning (ML) framework for the development of a MMS, which can be a better approach than previously developed approaches. The ML framework uses a Python-based solution methodology in conjunction with geo-spatial modeling, which appears more attractive and efficient in working directly with GIS maps and databases. With respect to application, the attention is focused on African countries using Kenya as a case study example. A recent report on state of Kenyan roads found over 35 percent of Kenyan roads to be still in poor condition even though a comparison of the condition of the roads between 2003 and 2018 showed a successive improvement in road condition over the years. Poor road condition affects mobility and, in turn affects the country’s economy. We adopt a Markov Decision Process to predict the maintenance actions to be undertaken for the Kenyan road network in order to keep an acceptable level of service quality over a specified planning horizon. A budget can then be estimated based on the cost of maintenance actions. A case study using Geographic Information System maps and databases demonstrates the effectiveness of the approach. The result shows that an MMS for Kenyan roads can help forecast the maintenance activities to be undertaken over a planning horizon. For more realistic practical applications, using some of our previous works as a guide, an algorithm to decide on the level of deterioration over time can be developed in future works which could consider factors like weather, vehicle mix, and traffic load.
    VL  - 10
    IS  - 4
    ER  - 

    Copy | Download

Author Information
  • Advancement Strategy Consulting, Columbia, MD, USA

  • Advancement Strategy Consulting, Columbia, MD, USA

  • Sections